Simulating Spring Final Warmings in Historical Runs of CMIP6 Models

The dynamics of the stratosphere play a significant role in medium and long-term weather forecasting and long-term trends in the stratosphere play a key role in the tropospheric climate system through a variety of direct and indirect mechanisms. As such, the importance of a well-simulated stratosphere in accurately modeling the Earth System has become well established in recent years. To assess the veracity of stratospheric simulations in current models, a suite of Historical runs from different members of the CMIP6 experiment are analyzed. For each model, the climatological annual cycle for the zonal winds and air temperature in the lower stratosphere and in the lower troposphere are calculated and compared with observations in the latter half of the 20^th century until present. This comparison shows the ability of different models to reproduce the observed seasonal cycle and also allows an assessment of the correlation between having a correctly simulated stratosphere and troposphere. Any model with significant biases in either wind or temperature in either the stratosphere or troposphere is discarded. Employing the remaining models, the author performs a composite study of the stratospheric final warmings (SFWs) and the boreal extratropical circulation. SFW events have been shown to provide a strong organizing influence upon the large-scale circulation of the stratosphere and troposphere during the period of spring onset. In contrast to the climatological seasonal cycle, SFW events noticeably sharpen the annual weakening of high-latitude circumpolar westerlies in both the stratosphere and troposphere. For each model, the average date of spring onset as well as other descriptive statistics are calculated as well as the composite evolution of zonal wind anomalies and temperatures as they propagate from the stratosphere down to the surface. These composite patterns are then compared with the canonical evolution as based on observations. These results are binned separately based on the stratospheric resolution of the model (so called high-top and low-top models) as well as the strength of the climatological wintertime polar vortex to identify biases present in different classes of the models.